8/7/2019 Saarna

http://slidepdf.com/reader/full/saarna 1/3

4th International DAAAM Conference“INDUSTRIAL ENGINEERING – INNOVATION AS COMPETITIVE EDGE FOR SME”29 - 30th April 2004, Tallinn, Estonia

RAIL AND RAIL WELD TESTING

M. Saarna, A. LaansooMechanical Testing and Calibration Laboratory of Department of Materials Engineering

Tallinn University of TechnologyE-mail: mklab@edu.ttu.ee

Abstract: Quality of rails and rail welds plays a significant role

in railway safety and general functioning. Testing of rails and 

rail weld plays a major part in quality control for rails and rail

welds.

Testing of rails is regulated by relevant standards, regulations

and legislations. The testing of rails and rail welds is directly

linked to the quality assurance programs of the railway

entrepreneurs or the rail welder company.

Key words: rail; testing; standards; rail welds.

1. INTRODUCTION

Quality of rails and rail welds plays a significant role in railway

safety and general functioning. Testing of rails and rail weld  plays a major part in quality control for rails and rail welds.

Railway safety is also closely linked to the quality of joiningthe rails by welding, which is directly linked to the quality

assurance programs of the railway entrepreneurs or the railwelder company. Testing of rails and rail welds are regulated

 by relevant standards, regulations and legislations.Choosing the right test parameters is vital for adequate test

results. The standards are not always the only source when

choosing the test parameters, relevant test standards can besupported by research results to choose the right test parameters.With the arrival of the new USA locomotives, which have a

greater load per axle than the old Russian ones, and due to theincreased rail heavy haul traffic the rails are subjected to even

greater stresses and are more likely to break as a result.The costs of a rail accident are high and may have long timeeffect on the credibility of the railway entrepreneur.

The main objective of the study is to review the mainrequirements for rails and rail testing, which can be obtained

from the relevant standards, which give the reference values for testing or from individual specific requirements for a country.

2.RAILS

Rail main function is to withstand the loads and guide the rail

cars. Due to the harsh working environment rail has towithstand corrosion, heat changes, and withstand cyclic loads.

All this leads to rail wear and to a possible rail failure.Also the rail has to be machinable and weldable. To reduce the

rail wear some of the rails are hardened either only from thehead or the whole rail (EVS EN 13674-1; GOST R 51685-

2000)Rail steel is commonly carbon- manganese steel with mostly

 pearlitic microstructure. The methods used to make the steel for rails vary between the oxygen furnaces and more moderns

electric-arc furnaces with vacuum degassing to remove

hydrogen and oxygen and other gases from the molten steelfrom. The hydrogen is most likely to cause cracks in the rail

head (Sperry Rail Service, 1999; D.F. Cannon et.al., 2003).For making the blooms either ingot or continues casting

methods are used.

The chemical composition, tensile strength and hardness valuesdiffer with the different grades of rails. All in all there are 7

grades including non heat treated carbon manganese, nonheated alloy (1%Cr), heat treated carbon manganese and lowalloy steels in the European standard and 11 grades in the

Russian standard. In the table 1 is shown the chemicalcomposition of grade R350HT and Э76T steels that are mainly

used to for new railway lines in Estonia. The rails profiles used

are in Estonia are European 60E1 and Russian P65, P50 andP43.

Table 1. Chemical composition of grade R350HT and Э76T

C, % Si, % Mn, % P, % S, %

R350HT

0.72…0.80 0.15…. 0.58 0.70…1.20 0.020 0.008…0.030

Э76T

0.71…0.82 0.25…0.45 0.75…1.05 0.025 0.030

2.1 Rail steel defects

Rail failures can be divided into three main groups:

•  the rail manufacturing defects e.g. hydrogen cracks;

•  defects due to the inappropriate handling, installation anduse;

•  defects caused by the exhaustion of the rail steels inherit

to resistance to fatigue damage (D.F. Cannon et.al.,2003).

Due to the harsh working environment rails are susceptible tofailure. In addition to bending and shear stresses rails have to

withstand dynamic loads and contact, thermal and residualstresses. The contact stresses between the wheel and rail are

high, up to 4000 MPa routinely and up to 1500 MPa with axleloads up to 37 t and contact are between the wheel and rail

around 1 cm2 (D.F. Cannon et.al., 2003;Telliskivi, Tet.al.2000).

The flattened wheels and also irregularities in the surface

condition of the rails mostly cause dynamic stresses.Thermal stresses are cause by the change in the weather temperature, in Estonia that could be in the range of -25ºC…+30ºC.

Residual stresses are introduced into the rail during thestraightening process during manufacturing.

Majority of rail failures are due to the transverse defects caused  by the either rail inner defects such as inclusions or fatiguecracks (D.F. Cannon et.al., 2003; Skinner,D.H.;Judd P.A.,

1981).The main reason for forming inner defects is hydrogenand oxygen. The hydrogen can hydrogen cracks within the steel

that can lead to formation of kidney shape transverse defectsupon traffic. Vacuum degassing and control cooling is used toextract gases (Sperry Rail Service, 1999).

For determining the sulphur segregation sulphur print method isused. Sulphur is the main cause of non-metallic inclusions in

rail steel that could form a transverse crack, especially whenclustered together. The critical size of the inclusion was shown

217

8/7/2019 Saarna

http://slidepdf.com/reader/full/saarna 2/3

to be about 80 µm and more for initiating a transverse fatigue

defect. For forming a transverse defect larger inclusion

diameter is needed (Skinner,D.H.;Judd P.A., 1981).The microstructure of the rail steel is supposed to be pearlitic,no martensite, bainite or grain boundary cementite is allowed.

Such high stresses and cyclic loading demand the railsteel to have adequate mechanical properties e.g. tensile

strength and elongation, fracture toughness, hardness, fatigue

crack growth rate and also impact strength.Requirements for rail steel are given in the European standardEVS EN 13674-1.

2.2 Rail testing according to EVS EN 13674-1

Rail steel testing is usually done as a part of rail manufacturing  process following the quality management system of themanufacturer.

To assess the lifetime of rails the fatigue crack growth ratesmust be known, to predict the crack propagation, the Paris law

da/dN = C(∆K)n is used to characterize the fatigue crack growthrates. Fatigue crack growth rate tests are carried out according

to the requirements of the BS 6835-1, a three-point bend andsingle edge notch test piece shall be used, with stress ratio

should be kept at 0.5. (L F M da Silva et.al., 2003)Fatigue testing are carried out according to ISO 1099

with symmetrical cycle about the initial zero load. Thequalifying criterion is for total strain amplitude of 0,00135 is

minimum of 5x106 cycles.Hardness tests are conducted on the running surface

of the rail and on the cross section of the rail for the heat-treated rails. The preferred hardness testing method is Brinellhardness according to EN ISO 6506-1 with test load of 1.839

kN and 2.5 mm diameter tungsten carbide ball indentor.Tensile tests are carried out on round specimens with

diameter of 10 mm and gauge length of 50 mm from therailhead crown. Tests are carried out according to EVS EN10002-1

Fracture toughness (K Ic), the tests are carried outaccording to ASTM E399 with specimen thickness of 25 mm

and width 40, 45 or 50 mm. The test temperature is -20ºC ±2ºC.

Chemical composition tests are to determine if the

steels conform to the requirements of EVS EN 13674-1.Impact testing is not foreseen in the new European

rail standard, but it does give information about the degree of toughness in lower temperatures. (Kecskes, S et.al., 1996).

The European standard also gives requirements for rail profiles,dimensions, straightness, surface flatness and twist.

3.WELDING METHODS

The two main welding methods used for joining rails are:thermite welding (TW) and flash welding (FBW). FBW is used

to join the short about 25 m rails to about 150 m rails which arethen shipped to the site and joined together by the TW to produce continuous rails.

Out of the two welding methods used the thermite welds aremore likely to fail due to their microstructure and cast defects

similarly to cast steel. About 17% of rail failures occur on TWin comparison 7.9% in FBW (D.F. Cannon et.al., 2003).

The greater number of failed TW welds is due to themicrostructure and macrostructure e.g. sulphur segregation

formed and also the chemical composition of the weld isdifferent from the rail the carbon content is smaller but the Mn-,

Si-content higher and also there is some Al present. Themicrostructure of the FBW weld is mostly pearlitic with some

ferrite and the TW the bead microstructure is bainitic and in theheat affected zone coarse-grained pearlite. As a result weld

  bead can be harder or softer then the rail material and causeuneven wear in that area, whish leads to the increase of noise

and stresses in rail. With the FBW welding the hardness changein the heat affected zone is not significant. The thermite weld

heat influence zone is wider then in the case of the FBW.The temperature of the TW welding process is higher then the

FBW welding about 2300-2400°C and 4000-5000°Caccordingly. (Kecskes, A; Kiss, Saba, 1996).The mechanical properties of the TW welds can be improved

  by a heat-treating. The main influence of the heat-treating is

making the coarse ferritic-pearlitic microstructure into muchfiner structure but the hardness of the heat-affected zonedecreases which in not favourable. The heat treatment for a

UIC 860 S49 type of rail was suggested to be a heating theweld up to 840 ºC holding the temperature for 45 minutes andair cooling (Ilic, N et.al., 1999).

3.1 Rail weld tests

The first test after the welding is visual examination of the weld

for visible defects such as geometry, and is usually carried out by the welding personnel on sight. The tests carried out in the

laboratory are hardness test, slow bend test, fracture surfaceexamination, micro- and macro structure examination.

Test methods for both TW and FBW welds are similar and the

requirements for thermite welding process are given in pr European standards EN 14730-1 Railway applications. Track.Aluminothermic welding of rails. Part 1: Approval of welding

 processes.The requirements for flash butt welding process are given in pr EN 14587-1 Railway applications. Track. Flash butt welding of 

rails. Part 1. New 220 and 260 grade rails in a fixed plant, pr EN 14587-2 Railway applications. Track. Flash butt welding of 

rails. Part 1. New 260Mn and 350 HT grade rails in a fixed plant.

3.1.2 Rail weld hardness tests

Hardness tests are to show the distribution if hardness in rail

  base material, weld and HAZ. The hardness test is the easiestway to estimate the quality of the rail weld in the first

approximation. By it’s own it cannot be used to determine thequality of a weld and usually it is complementary to other test

methods discussed here. The hardness numbers are to illustratethe type microstructure of the weld and HAZ.

According pr EVS EN 13674-1hardness tests include VickersHV30 to determine the distribution of hardness values of the

weld and HAZ and HBW 10/3000/15 to determine the weldhardness on the centre line of the head crown.

According to ANSI/AWS D15.2-94 a Rockwell hardnesssuggested to determine the hardness traverse on a vertical

central longitudinal section plane, and Brinell 10/3000 on thehead, it seems that the Vickers method gives better resolution

of the hardness traverse due to the fact that the indentations can  be made more close to each other then the Brinell methodwould allow.

3.1.3 Rail weld bend tests

Rail bend test is carried out to determine the quality of the railweld by the means of examining results of the static bending

test.Rail is either bent to some qualifying criterion load or to break.

After the rail fracture the fracture surfaces are examined for macrostructure e.g. visual defects such as inclusions, lack of 

weld.For both thermite and flash butt weld the test scheme is the

same according to pr EN 14730-1 and pr EN 14587-2, 3-point

 bend test (fig.1).The maximum loading speed is 1 mm/s. The parametersmeasured are load at break and maximum deflection.

218

8/7/2019 Saarna

http://slidepdf.com/reader/full/saarna 3/3

 

Fig.1 Bend testing scheme

According to ANSI/AWS D15.2-94 the span length is 1200 mmand loading instead of 3-point bend scheme a 4-point loading

scheme is suggested to be used (fig.2).The load is applied at two loading points 300 mm apart. Load is

applied until the rail fractures or deflects 100 mm, which ever comes first.

Fig.2. 4-point bend test scheme suggested by the ANSI/AWSD15.2-94

3.1.4 Rail weld fatigue tests

Rail fatigue testing is carried out, accordingly to pr EN 14730-1in the case of TW and pr EN 14587-1 in case of FBW, on a rail

weld with a 4-point type of bend test schemeThe qualifying criterion for cyclic load is, with ratio min/maxload ratio of 0.1 and frequency of 10 Hz, a minimum of 5x106 

cycles.According to ANSI/AWS D15.2-94 the rail weld fatigue testing

is suggested to be carried out by rolling load test scheme. Testspeed is 60 cycles per minute (1Hz) for total of 2 million cycles

or to the failure which ever comes first (fig.3).

Fig. 3 Rolling load test scheme suggested by the ANSI/AWSD15.2-94

5. CONCLUSION

1000 mm 

The main rail mechanical testing methods for rails and railwelds were reviewed.

Rail testing is closely linked to the relevant standards andtesting methods. When choosing the test parameters for some of 

the standard test methods, it is advisable to consult the relevant

articles for the preferred test parameters and guidelines.The increase of heavy haul rail traffic in Estonia has raised theneed for rail testing. The testing is carried out according to the

relevant European standards.In addition to rail testing monitoring the roundness of the rail-car and locomotives wheels is important to reduce the dynamic

shock loading of rails. Also monitoring the state of theembankment is important to reduce the possible affect of the

additional stresses to the rails.

6. REFERENCES

ANSI/AWS D15.2-94 Recommended Practices for the Welding

of Rails and Related Rail Components for use by Rail Vehicles.An American standard.

D.F. Cannon, K.-O. Edel, S.L. Grassie and K. Sawley.(2003).Fatigue Fract. Engng. Mater. Struct , 26, 865-887. Blackwell

Publishing Ltd.

  pr EN 14730-1 Railway applications. Track. Aluminothermicwelding of rails. Part 1: Approval of welding processes.

EVS EN 13674-1 Railway applications - Track - Rail - Part 1:Vignole railway rails 46 kg/m and above.

GOST R 51685-2000 Railway rails. General specifications.

Russian standard.

Ilic, N; T. Jovanovic, M; Todorovic, M; Trtanj, M andSaponjic, P. Microstructual and Mechanical Characterization of Post weld Heat-Treated Thermite weld in Rails. Materials

Characterization 43:243-250, 1999.

Kecskes, S; Kiss, Csaba, (1996), Cause of the formed valley atthe welding bead noticed by the supervisory staff of the track 

maintenance, 2-nd Mini. Conf. Of contact Mechanics and Wear of Rail/Wheel Systems, Budapest, July 29-31, 1996.

Kecskes, S; Kiss, Csaba; Szeman, L. (1996). Comparison of 

home made rails with import rails, and determination of  probable rail lifetime by tensile test of notched specimen.

 pr EN 14587-1 Railway applications. Track. Flash butt weldingof rails. Part 1. New 260Mn and 350 HT grade rails in a fixed plant.

Skinner,D.H.;Judd P.A., (1981), A Metallographic study of fatigue defects in rails. Metals in mining: Proc. Of the 34-th

Annual Conference, 68-72.

Sperry Rail Service (1999) Rail Defect Manual. Danbury,

Connecticut, USA.

Telliskivi T.; Olofsson U., (2000). A tool and a method for FEanalysis of wheel and rail interaction, 10th internationalANSYS conference and exhibition, Pittsburgh August 28-30.

219